Nanocomposites of Nitrogen-Doped Graphene Oxide and Manganese Oxide for Photodynamic Therapy and Magnetic Resonance Imaging

Cancer is a leading cause of death worldwide. Conventional methods of cancer treatment, including chemotherapy and radiotherapy, are associated with multiple side effects. Recently, photodynamic therapy (PDT) has emerged as an effective therapeutic modality for cancer treatment without adversely affecting normal tissue. In this study, we synthesized nitrogen doped graphene (NDG) and conjugated it with Mn3O4 nanoparticles to produce NDG-Mn3O4 nanocomposite with the aim of testing its bimodal performance including PDT and magnetic resonance imaging (MRI). We did not use any linker or binder for conjugation between NDG and Mn3O4, rather they were anchored by a milling process. The results of cell viability analysis showed that NDG-Mn3O4 nanocomposites caused significant cell death under laser irradiation, while control and Mn3O4 nanoparticles showed negligible cell death. We observed increased generation of singlet oxygen after exposure of NDG-Mn3O4 nanocomposites, which was directly proportional to the duration of laser irradiation. The results of MRI showed concentration dependent enhancement of signal intensity with an increasing concentration of NDG-Mn3O4 nanocomposites. In conclusion, NDG-Mn3O4 nanocomposites did not cause any cytotoxicity under physiological conditions. However, they produced significant and dose-dependent cytotoxicity in cancer cells after laser irradiation. NDG-Mn3O4 nanocomposites also exhibited concentration-dependent MRI contrast property, suggesting their possible application for cancer imaging. Further studies are warranted to test the theranostic potential of NDG-Mn3O4 nanocomposites using animal models of cancer.


Introduction
Cancer is one of the deadliest and costliest diseases and is the second leading cause of death worldwide. Photodynamic therapy (PDT) is a promising treatment modality for cancer with minimal side effects and is expected to replace traditional chemotherapy, which is associated with numerous adverse effects. PDT involves combination of light and a photosensitizer (PS), which is activated by absorption of light of a specific wavelength, causing the generation of potentially toxic reactive oxygen species (ROS) that induce a cascade of intracellular molecular events resulting in targeted tissue damage [1,2]. Sun et al. have reviewed the application of metal-based nanoparticles (NPs) for PDT of cancer [3]. Metal oxide-based nanomaterials have also significantly impacted the landscape of healthcare, including in the areas of diagnosis and therapeutic applications [4]. Metal oxides have studied the cytotoxicity of these nanocomposites and evaluated their efficiency in PD and MRI using in-vitro models. We also tested the ability of NDG-Mn3O4 nanocomposi to generate ROS under laser irradiation as a potential mechanism of their toxicity in canc cells. This is probably the first study reporting the nanocomposite of NDG and Mn3 nanoparticles.

Results
The results of high resolution transmission electron microscopy (HRTEM) display the existence of spherical shaped Mn3O4 nanoparticles on the surface of NDG within t range of 5-15 nm (Figure 1). The Mn3O4 NPs are well distributed on the surface of ND as the magnified image indicates the shape and crystallinity of these NPs. The Mn3O4 N are not bonded covalently but are held by physisorption on the NDG surface by Vand Waals interactions. The elemental composition of NDG-Mn3O4 nanocomposite, analyz by energy-dispersive X-ray spectroscopy, showed intense signals at 0.65, 5.88, and 6. keV strongly suggesting that 'Mn' was the major element, which has an optical absorpti in this range owing to the surface plasmon resonance (SPR). Other signals that were fou in the range of 0.0-0.5 keV signified the absorption of carbon, nitrogen and oxygen, co firming the formation of NDG-Mn3O4 nanocomposite. The average particle size of t NDG-Mn3O4 nanocomposite was found to be 10 ± 1.7 nm (Figure 1).   [30]. Notably, the sharp diffraction peaks point toward the highly crystalline and well-disperse nature of nanoparticles which clearly matched with the Hausmannite crystal phase [31]. On the other hand, the XRD pattern of NDG-Mn 3 O 4 nanocomposite showed the appearance of a broad peak at~22.4 • (002) (Figure 2b) that confirmed the reduction of graphene oxide and formation of NDG [32]. Furthermore, there is no broadening or shift of the (002) peak, proving that there is no change in the interlayer spacing of graphene after nitrogen-doping. No significant change in the full width at half-maximum (FWHM) of the (002) diffraction peak indicates the similar crystallite size before and after nitrogen doping [33]. In case of the composite, the XRD pattern of which is shown in Figure 2c, characteristic diffraction peaks of both Mn 3 O 4 and N-doped graphene are present, which clearly indicate the formation of hybrid material. FT-IR spectra of Mn3O4 NPs displayed the characteristic peak of Mn-O, stretching mode in the range of 624 cm −1 while the vibrational frequency associated to the Mn-O distortion vibration poisoned at 525 cm −1 (Figure 3a). The characteristic narrow and broad bands located at 3420 and 1600 cm −1 were related to the hydroxyl (-OH) groups absorbed by the samples or potassium bromide. FT-IR spectra of NDG are shown in Figure 3b. FT-IR spectra of NDG-Mn3O4 displayed the graphene oxide intense bands for C=C stretching (~1630 cm −1 ), C-O-C stretching (~1209 cm −1 ), C-O stretching (~1050 cm −1 ). The nitrogen doping in the sample was confirmed by the presence of two characteristic peaks at ~1325  (Figure 3a). The characteristic narrow and broad bands located at 3420 and 1600 cm −1 were related to the hydroxyl (-OH) groups absorbed by the samples or potassium bromide. FT-IR spectra of NDG are shown in Figure 3b. FT-IR spectra of NDG-Mn 3 O 4 displayed the graphene oxide intense bands for C=C stretching (~1630 cm −1 ), C-O-C stretching (~1209 cm −1 ), C-O stretching (~1050 cm −1 ). The nitrogen doping in the sample was confirmed by the presence of two characteristic peaks at~1325 and~1570 cm −1 , which were attributed to the stretching of the C-N bond from the secondary aromatic amine, which pointed toward bonding between carbon and nitrogen including the existence of other absorption bands of 'Mn' at 624 and 525 cm −1 clearly indicating the formation of HRG-Mn 3 O 4 nanocomposite (Figure 3c).   The results of cell viability analysis using MTT assay showed that exposure of Mn 3 O 4 and NDG-Mn 3 O 4 in the concentration range of 6.25-100 µg/mL did not cause any cytotoxicity ( Figure 4). However, NDG-Mn 3 O 4 nanocomposites displayed significant cells death under laser irradiation for 5 min, while PBS (control) and Mn 3 O 4 NPs showed negligible cell death ( Figure 5). Almost 100% cells were viable when treated with PBS whereas 41% for cancer cells survived after the treatment of 100 µg/mL concentration of NDG-Mn 3 O 4 nanocomposites along with 5 min of laser irradiation. The effect of NDG-Mn 3 O 4 nanocomposites on the cytotoxicity of A549 cells was concentration-dependent and only the concentrations of 25 µg/mL and above were found to be effective in killing the cells under laser irradiation ( Figure 5). The results of cell viability analysis using MTT assay showed that exposure of Mn3O4 and NDG-Mn3O4 in the concentration range of 6.25-100 μg/mL did not cause any cytotoxicity ( Figure 4). However, NDG-Mn3O4 nanocomposites displayed significant cells death under laser irradiation for 5 min, while PBS (control) and Mn3O4 NPs showed negligible cell death ( Figure 5). Almost 100% cells were viable when treated with PBS whereas 41% for cancer cells survived after the treatment of 100 μg/mL concentration of NDG-Mn3O4 nanocomposites along with 5 min of laser irradiation. The effect of NDG-Mn3O4 nanocomposites on the cytotoxicity of A549 cells was concentration-dependent and only the concentrations of 25 μg/mL and above were found to be effective in killing the cells under laser irradiation ( Figure 5).   The results of cell viability analysis using MTT assay showed that exposure of Mn3O4 and NDG-Mn3O4 in the concentration range of 6.25-100 μg/mL did not cause any cytotoxicity ( Figure 4). However, NDG-Mn3O4 nanocomposites displayed significant cells death under laser irradiation for 5 min, while PBS (control) and Mn3O4 NPs showed negligible cell death ( Figure 5). Almost 100% cells were viable when treated with PBS whereas 41% for cancer cells survived after the treatment of 100 μg/mL concentration of NDG-Mn3O4 nanocomposites along with 5 min of laser irradiation. The effect of NDG-Mn3O4 nanocomposites on the cytotoxicity of A549 cells was concentration-dependent and only the concentrations of 25 μg/mL and above were found to be effective in killing the cells under laser irradiation ( Figure 5).   The results of in-vitro photodynamic therapy are shown in Figure 6. Without laser irradiation, none of the treatments including PBS, Mn3O4, or NDG-Mn3O4 caused any cellular damage as almost all the cells appeared green. After 5 min laser irradiation, NDG-Mn3O4 nanocomposites killed 68% of the cancer cells (shown as red dots) whereas the treatments of PBS and Mn3O4 did not cause any significant cellular damage under laser irradiation ( Figure 6). To evaluate the 1 O2 generation from NDG-Mn3O4 nanocomposites under laser irradiation, we measured the absorbance of 1,3-diphenylisobenzofuran (DPBF) after laser irradiation (670 nm, 0.1 W/cm 2 ) at different time points (Figure 7). The DPBF absorbance decreased with increasing the laser irradiation time, indicating the generation of singlet oxygen from NDG-Mn3O4 nanocomposites is directly proportional to the duration of laser irradiation (Figure 7).  For testing the effectiveness of NDG-Mn3O4 nanocomposites toward diagnostic standpoint, we investigated whether these nanoparticles have MRI contrast properties or not. Various concentrations of nanoparticles were subjected to imaging by 3T MRI scanner. The result demonstrated a concentration dependent enhancement of signal intensity with increasing concentration of NDG-Mn3O4 nanocomposites. The r1 value was found to be 0.09 mM −1 s −1 (Figure 8). For testing the effectiveness of NDG-Mn3O4 nanocomposites toward diagnostic standpoint, we investigated whether these nanoparticles have MRI contrast properties or not. Various concentrations of nanoparticles were subjected to imaging by 3T MRI scanner. The result demonstrated a concentration dependent enhancement of signal intensity with increasing concentration of NDG-Mn3O4 nanocomposites. The r1 value was found to be 0.09 mM −1 s −1 (Figure 8).

Discussion
In this study, we conjugated nitrogen doped GO with Mn 3 O 4 nanoparticles to explore the PDT and MRI potentials of these entities. Nafiujjaman et al., [34] developed a ternary hybrid probe as a dual imaging-guided PDT agent, which consisted of Mn 3 O 4 and graphene quantum dots (GQD) linked by polydopamine. Conjugation of dextran (DEX) with GO has been shown to remove agglomeration and improve the stability of GO in biological solutions [25]. Kim et al., [35] have shown that DEX coated GO (DEX-GO) nanoparticles are highly biocompatible, maintaining a high cellular viability (>80%) at high concentrations (450 µg/mL). The hybrid DEX-GO nanoparticles also showed significantly higher (18.6 fold) optical density in the NIR region as compared to GO alone, suggesting the application of DEX-GO for PDT [25]. Chen et al., [36] coated iron oxide on DEX-GO nanoparticles and observed that the transformed nanoparticles not only showed better MRI contrast property but also exhibited negligible toxicity.
The results of cell viability analysis showed that NDG-Mn 3 O 4 nanocomposites did not cause any cytotoxicity unless activated by laser irradiation that resulted in concentration dependent cytotoxicity in lung cancer cells ( Figure 5). These biochemical findings were supported by fluorescence microscopy observations, suggesting that NDG-Mn 3 O 4 nanocomposites initiate cytotoxic properties only under laser irradiation ( Figure 6). We used 670 nm laser in order to keep the wavelength within the red optical window (620-750 nm). Visible red radiation is able to activate photosensitizers in deep tumors without causing phototoxicity to normal tissue [37,38]. In our study, NDG-Mn 3 O 4 nanocomposites killed 68% of cancer cells which is more effective than GQD-PDA-Mn 3 O 4 nanoparticles (51% cell death) as reported earlier [34]. The mechanism of laser-induced toxicity of NDG-Mn 3 O 4 nanocomposites can be multifactorial. Nafiujjaman et al. [34] have reported that laser irradiation during PDT triggered the disruption of cellular membranes resulting in a higher cellular uptake of the GQD-PDA-Mn 3 O 4 nanoparticles compared to graphene quantum dots. This selective transport across the cell membrane might have been influenced by the size, shape and surface chemistry of nanoparticles.
GO-based nanomaterials possess improved tumor-passive targeting effect and comparatively higher tumor uptake than CNTs due to the enhanced permeability and retention effects, attributed to the peculiar two-dimensional structure and small lateral size of GO [39]. By virtue of its unique optical properties, GO can be used for live cell imaging (due to near infrared photoluminescence) [40][41][42] as well as for PDT (due to free radical generation by optical absorption) [43][44][45]. Because of the minimal cellular autofluorescence in the NIR region, the chances of interference during imaging are also reduced [41]. Thus, intracellular tracking of GO-based nanomaterials can be performed without conjugation with fluorescence dyes, due to the photoluminescence property of GO.
The mechanism of PDT is complex and not fully understood. It has been suggested that singlet oxygen ( 1 O 2 ) plays a key role in PDT, which is formed in the molecules of lipids and proteins of cell membranes and intracellular organelles when exposed to the quantum of light [46]. Singlet oxygen is cytotoxic for living cells due to its strong oxidizing property [47]. Our results showed that NDG-Mn 3 O 4 nanocomposites caused 1 O 2 generation under laser irradiation in a time-dependent manner and longer exposure to laser irradiation produced excessive ROS generation (Figure 7). Because uncontrolled generation of ROS is deleterious to normal cells [48], we selected 5 min laser irradiation in further experiments. Sustained elevated levels of ROS leads to irreparable cellular damage through oxidation of nucleic acids, lipids, and proteins, ultimately resulting in cell death through apoptosis or necrosis. There are different scales of oxidative stress ranging from physiological oxidative stress to excessive and toxic oxidative burden [49]. Photosensitizers with near-infrared (NIR) fluorescence as well as efficient ROS generation ability have been used for precise diagnosis and simultaneous treatment of cancer [50]. Liu et al., [51] designed a novel multi-functional nanosystem in which cisplatin was loaded in MnO 2 -doped GO and functionalized with a photosensitizer (Ce6). The nanosystem was equipped with intelligent functions including: (1) decomposition of H 2 O 2 into oxygen to relieve the tumor hypoxia; (2) depletion of glutathione (GSH) in tumor cells; and (3) initiation of Mn 2+ medicated Fenton-like reaction to generate ROS, all of which contributed to the enhanced anti-tumor efficacy of the nanomaterial [51]. Nonlinear absorption of two relatively low-energy photons of NIR light is associated with the emission of high-energy visible light that can sensitize oxygen to produce cytotoxic ROS including singlet oxygen which can kill cancer cells [52]. Lu et al., [53] attributed cell death caused by mitochondria-mediated apoptosis of MCF-7 breast cancer cells exposed to low concentration (1.6 µg/mL) of Ir(tiq) 2 ppy NPs under white light irradiation at quite low intensity (5 mW cm −2 ) to excessive generation of ROS under light irradiation. Hou et al., [54] developed a multifunctional nanoplatform to enhance PDT efficiency by increasing the generation of ROS in tumor cells through Fenton reaction and reducing the distance between ROS and target site by targeting the mitochondria.
The results of MRI demonstrated a concentration dependent enhancement of signal intensity with increasing concentration of NDG-Mn 3 O 4 nanocomposites (Figure 8). Although MRI is the best imaging technique for detecting soft tissue, the long relaxation time of water protons led to weak differences between tissues, resulting in poor image depiction between normal and malignant tissue [55]. However, the use of contrast agents (CAs) significantly enhanced the quality of MRI images and therefore the sensitivity of MRI-based clinical diagnosis [56]. Gadolinium (Gd)-based T1 contrast agents have been commonly used in clinical practice [57], however, they have the drawbacks of short blood circulation time and nephrotoxicity [58]. In recent years, due to their good biocompatibility, relatively high magnetization spin and rapid water proton exchange rate, manganese oxide nanoparticles have been developed as T1 contrast agents that have shown significant potential for detection and diagnosis of cancer [9,28,29].

Preparation of Mn 3 O 4 Nanoparticles
Manganese (II) acetylacetonate was dissolved in oleylamine (molar ratio of manganese (II) acetylacetonate: oleylamine = 1:25) and the mixture was heated at 160 • C for 10 h under a nitrogen cover. The resulting product was cooled to room temperature to form a brownish suspension, which was centrifuged at 9000 rpm for 15 min and the supernatant was removed to obtain a brown residue. The precipitate was washed multiple times with ethanol to acquire pure Mn 3 O 4 nanoparticles, which were dried under vacuum before use [34].

Preparation of Nitrogen-Doped Graphene Oxide (NDG)
Initially, graphite oxide (GO) was synthesized from graphite powder using a modified Hummers method [59,60]. Briefly, graphite powder (0.5 g) and NaNO 3 (0.5 g) were added to 23 mL of H 2 SO 4 and the mixture was stirred for 10 min in an ice bath. Subsequently, KMnO 4 (3 g) was slowly added and after proper mixing, the ice bath was replaced with water bath (35 • C) for 1 h, resulting in the formation of a thick paste. Thereafter, 40 mL of deionized water was added, and the mixture was stirred for 30 min at 90 • C. Finally, 100 mL of deionized water was added, followed by the slow addition of 3 mL of H 2 O 2 . The mixture was allowed to cool, filtered and washed with deionized water. The resulting thick brown paste was dispersed in water and centrifuged at 1000 rpm for 2 min. This step was repeated 4-5 times, until all unsettled particles were removed. The resultant paste was dispersed in water with mild sonication to obtain a suspension of graphene oxide (GO). For nitrogen doping, the resulting suspension was taken in a round bottom flask, to which 4 mL of NH 4 OH and 4 mL hydrazine hydrate were added simultaneously. The mixture was stirred for a few minutes, and the flask (equipped with cooling condenser) was put in a water bath controlled at 90 • C for 3 h. The product was collected after been filtered through micropore filters (Whatman filter paper, pore size-20 µm, W&R Balston Limited, Maidstone, Kent, UK), washed by deionized water and freeze-dried.

Preparation of Nanocomposites of NDG and Mn 3 O 4 (NDG-Mn 3 O 4 )
Equal amounts of Mn 3 O 4 nanoparticles and NDG were milled using a Fritsch Pulverisette P7 planetary ball mill (Idar-Oberstein, Germany). The nanomaterials powder and stainless steel balls (5 mm diameter) with the ball to powder weight ratio of 1:1 were introduced into the stainless steel container. The milling of the powder was performed for 16 h, with intermittent pausing of milling process at regular intervals.

Cell Viability Analysis
We used the 3-(4,5-dimethylthiazol-2-y1)-2,5-diphenyltetrazolium bromide (MTT) method for testing the cytotoxicity of Mn 3 O 4 and NDG-Mn 3 O 4 nanoparticles. A549 lung cancer cells were seeded into 96-well plate (4 × 10 4 cells per well) in RPMI and incubated at 37 • C for 4 h in a 5% CO 2 incubator. Different concentrations (6.25-100 µg/mL) of Mn 3 O 4 and NDG-Mn 3 O 4 nanoparticles were added to the 96-well plate. Phosphate buffer saline (PBS) was used as a control whereas triton-X100 was used as negative control. The cells were treated with a 670 nm laser irradiation at 0.1 W/cm 2 for 5 min and further incubated for 24 h. Aqueous solution of MTT (50 µL) was added to each well in the 96-well plate 4 h before the termination of 24 h incubation. The upper layer of the solution was discarded. The MTT solubilization solution, DMSO (100 µL) was added to each well to dissolve the formazan crystals by pipette stirring and then observed the absorbance at 590 nm, which was converted to cell viability using the following equation [34].

In-Vitro Photodynamic Therapy
We used fluorescence microscopy for morphological analysis of cancer cells following treatment with nanoparticles and laser irradiation. Fluorescein diacetate (FDA) and propidium iodide (PI) were used to visualize the live and dead cells, respectively. A549 cells (2 × 10 4 cells per well) were seeded in a 24 well plate and incubated at 37 • C for 24 h in an atmosphere of 5% CO 2 . Mn 3 O 4 and NDG-Mn 3 O 4 nanoparticles (50 µg/mL) were added to the wells and the plate was incubated for 4 h. After incubation, the cells were irradiated for 5 min with a 670 nm laser, followed by another incubation for 24 h. Both the dyes were added to wells and the plate was incubated for 5 min. Then, the cells were washed three times with PBS to remove excess dyes, and the fluorescence images were acquired by fluorescence microscope with 490 nm excitation and 525 nm emission wavelengths.

Analysis of Singlet Oxygen Generation
We used 1,3-diphenylisobenzofuran (DBPF) to detect singlet oxygen ( 1 O 2 ) generation by NDG-Mn 3 O 4 nanocomposites under 670 nm laser irradiation (0.1 W/cm 2 ). Fifty microliters of ethanolic solution of DPBF (1 mg/mL) were added to the nanocomposites solution under stirring and irradiated with laser for different time points. The absorbance of solution was measured by UV-Visible spectrophotometer. The decrease in absorbance at 426 nm indicated the degradation of DPBF in presence of 1 O 2 which was generated by laser-induced activation of NDG-Mn 3 O 4 nanocomposites.

MRI Relaxivity Analysis
A series of aqueous suspensions of NDG-Mn 3 O 4 nanoparticles (with Mn concentration from 0 to 1 mM) were prepared and imaged in 0.2 mL Eppendorf tubes using a 3T clinical MRI instrument (GE Signa Excite Twin-Speed, GE Healthcare, Milwaukee, WI, USA). The specific relaxivity (r 1 ) was calculated from linear curve generated from concentration of NDG-Mn 3 O 4 nanocomposites versus 1/T 1 (s −1 ).

Statistics
The data were analyzed by one-way analysis of variance (ANOVA) followed by Dunnett's test. p values < 0.05 were considered as statistically significant.

Conclusions
The NDG-Mn 3 O 4 nanocomposites did not cause any cytotoxicity under physiological conditions. However, they produced significant and dose-dependent cytotoxicity in cancer cells after 670 nm laser irradiation. The PDT potential of NDG-Mn 3 O 4 nanocomposites was attributed to excessive generation of ROS in exposed cells. NDG-Mn 3 O 4 nanocomposites also exhibited concentration-dependent MRI contrast property suggesting their possible application for cancer imaging. Further studies are warranted to test the theranostic potential of NDG-Mn 3 O 4 nanocomposites in animal models of cancer.